Assisting Scalable Diagnosis Automatically via CT Images in the Combat against COVID-19

Introductory paragraphThe pandemic of coronavirus Disease 2019 (COVID-19) caused enormous loss of life globally. 1-3 Case identification is critical. The reference method is using real-time reverse transcription PCR (rRT-PCR) assays, with limitations that may curb its prompt large-scale application. COVID-19 manifests with chest computed tomography (CT) abnormalities, some even before the onset of symptoms. We tested the hypothesis that application of deep learning (DL) to the 3D CT images could help identify COVID-19 infections. Using the data from 920 COVID-19 and 1,073 non-COVID-19 pneumonia patients, we developed a modified DenseNet-264 model, COVIDNet, to classify CT images to either class. When tested on an independent set of 233 COVID-19 and 289 non-COVID-19 patients. COVIDNet achieved an accuracy rate of 94.3% and an area under the curve (AUC) of 0.98. Application of DL to CT images may improve both the efficiency and capacity of case detection and long-term surveillance.

Diagnostic value of CT at early infection stage of thoracic and pulmonary paragonimiasis / 中华地方病学杂志

Objective To investigate the CT features of early infection stage of thoracic and pulmonary paragonimiasis. Methods Medical records of 56 patients with thoracic and pulmonary paragonimiasis from January 2010 to June 2017 were collected, and the patients were diagnosed and treated at Yongjia County People's Hospital, and the results of laboratory examination and CT imaging features were analyzed retrospectively. Results The absolute value of eosinophils in peripheral blood of 56 patients was (5.61 ± 3.18) × 109/L, and the percentage of eosinophils was (35.90 ± 19.16)％, all of which increased to varying degrees. Forty-two patients had different degrees of pleural effusion and 52 cases with lung lesions. Lung lesions demonstrated one or several kinds of foci at the same time, randomly distributed in the lung field, mostly located in the sub-pleural lung tissue. There were 12 cases with pulmonary ground glass shadow, 4 cases with peribronchitis, 31 cases with pulmonary invasive lesions and 28 cases with pulmonary nodular/strip shadow. The size of most nodules were 0.5 - 1.0 cm, accompanied with halo sign. Conclusions The CT features of early infection stage of thoracic and pulmonary paragonimiasis are diverse. The size of 0.5 - 1.0 cm lung nodules with halo sign has certain characteristics in the diagnosis of paragonimiasis. Peribronchitis, infiltrative lesions, pleural effusion and increased peripheral blood eosinophil percentage can suggest diagnosis.

Value of dynamic enhanced multi-slice spiral computed tomography in the differentiation of benign and malignant intraductal papillary mucinous neoplasm of pancreas / 中华胰腺病杂志

Objective To analyze the value of dynamic enhanced multi-slice spiral computed tomography (MSCT) combined with two-dimensional (2D) curved reconstruction technique in the differentiation of benign and malignant intraductal papillary mucinous neoplasm (IPMN) of pancreas,and compare with magnetic resonance cholangiopancreatography(MRCP).Methods MSCT and MRCP data of a total of 50 patients with IPMNs confirmed by pathology after surgery was retrospectively reviewed.The benign and malignant IPMNs were differentiated based on the presence of mural nodules,main pancreatic duct (MPD) ≥ 10 mm,septum thickness ≥2 mm,calcification,surrounding vascular infiltration,enlarged peripancreatic lymph nodules,distant metastatic lesions and maximal branch duct type IPMN lesions ≥30 mm shown in the images.The sensitivity,specificity and accuracy were calculated and the receiver-operating-characteristics (ROC) analysis were drawn.Area under the curve (AUC) was calculated.Results Mural nodules in MSCT had a sensitivity,specificity,and accuracy of 77.1％ (27/35),80.0％ (12/15) and 78.0％ (39/50) for diagnosing malignant IPMN,respectively;which in MRCP were 77.1％ (27/35),86.7％ (13/15),and 80.0％ (40/50) in comparison.When MPD diameter ≥10 mm was used for diagnose malignancy,MSCT and MRCP had the sensitivity,specificity,and accuracy of 96.3％ (26/27),81.8％ (9/11),92.1％ (35/38),and 96.3％ (26/27),90.9％ (10/11),94.7％ (36/38),respectively.For thick septum ≥2 mm,MSCT and MRCP had the sensitivity,specificity,and accuracy of 4.8.6％ (17/35),93.3％ (14/15),62.0％(31/50),and 51.4％ (18/35),93.3％ (14/15),64.0％ (32/50),respectively.Out of 50 cases,calcifications were detected on MSCT in 6 patients,and 5 of them were pathologically diagnosed as malignant IPMN.MRCP failed to identify calcifications in any of these lesions.For MSCT,the AUC of MPD diameter ≥ 10 mm,mural nodules and thick septum ≥ 2 mm were 0.973 (P =0.000),0.825 (P =0.002) and 0.704(P =0.051),respectively.For MRCP,the AUC of the three factors above were 0.976(P =0.000),0.825(P =0.002),0.722 (P =0.034),respectively.For the predicting of IPMN malignancy,MSCT had an overall sensitivity,specificity,and accuracy of 94.3％ (33/35),73.3％ (11/15) and 88.0％ (44/50),respectively;in comparison,MRCP had values of 94.3％ (33/35),80.0％ (12/15) and 90.0％ (45/50),respectively.Conclusions Presence of mural nodules,MPD ≥10 mm and thick septum ≥2 mm on MSCT combined with 2D curved reconstruction or MRCP have a high value for predicting the malignancy of IPMN.The values of MSCT and MRCP were basically consistent in the differentiation of benign and malignant IPMN.MSCT can be used as the preferred examination for diagnosing IPMN in the primary hospitals without MR equipment.

Analysis of genomic copy number variation for a Chinese patient with split hand/split foot malformation / 中华医学遗传学杂志

<p><b>OBJECTIVE</b>To employ single nucleotide polymorphisms (SNP) microarray to detect copy number variations (CNVs) for the diagnosis of disease and molecular classification.</p><p><b>METHODS</b>For a patient with split-hand/split-foot malformation, genome-wide copy number variants SNP microarray was applied. Tiny copy number variations were verified by real-time fluorescent quantitative PCR.</p><p><b>RESULTS</b>The results of SNP microarray has revealed that the patient has carried a 0.39 Mb duplication in 10q24.31-24.32 (102 955 122-103 348 688), which has encompassed genes including LBX1, BTRC and POLL. By real-time fluorescent quantitative PCR, duplicate area encompassing the pathogenic genes have been verified. The results for LBX1, BTRC, POLL genes were all consistent with the SNP microarray test. Moreover, a duplication was detected in exon 9 of FBXW4 gene which is in nearby.</p><p><b>CONCLUSION</b>SNP chips can efficiently identify tiny CNVs (< 1.0 Mb). In combination with real-time fluorescence quantitative PCR, this may provide valuable information for prenatal diagnosis.</p>

Breast Benign Diseases:Diagnosis with Molybdenum Target Mammography / 实用放射学杂志

Objective To discuss the diagnostic value of molybdenum target radiography in breast benign diseases.Methods 63 cases with breast benign diseases confirmed by operation and pathology underwent mammography,the clinical and X-ray signs were analysed retrospectively.Results 64 masses were found in the 63 cases,including 42 cases of fibroadenoma,10 cases of cyst,7 cases of duct papilloma,2 cases of hamartoma and 2 cases of hemangioma.All of the lesions,there were 32 with sharp border,22 with partly sharp border and 10 cases with fuzzy borders.12 lesions were accompanied by calcification.Conclusion The molybdenum target mammography has very important value in diagnosing and differentiating the benign breast diseases,it should be the first method of choice for female breast mass lesions.

Diagnostic Value of CR Mammography Signs for Breast Cancer / 实用放射学杂志

Objective To explore the imaging features of breast cancer by CR mammography,in order to improve the diagnostic level. Methods The CR mamographic features of 37 patients with breast cancer confirmed surgically and pathologically were analysed retrospectively. Results In the 37 cases of breast cancer,the following abnormal signs were found in CR mammograms: mass(n=29),without mass(n=8),calcification(n=15),abnormal vascularity(n=7),inverted nipple(n=6),thickened skin(n=3),sign of tower's tine(n=3),sign of funnel(n=1),focal structural disorder and small focus of increased density(n=3). Conclusion The masses and calcificationsare the main radiological features in breast cancer,CR mammography may show the characteristics of breast cancer,it is one of the mosteffective means in diagnosis of breast cancer.

X-ray Diagnosis of Neonatal Respiratory Distress Syndrome(NRDS) (An Analysis of 20 Cases) / 实用放射学杂志

Objective To improve the knowledge and diagnostic ability of imagiology of NRDS.Methods The clinical data and X-ray features in 20 cases with NRDS who had complete materials were analysed.Results According to X-ray manifestations all cases were divided into four stages:Ⅰ stage(n=3),Ⅱ stage (n=5),Ⅲ stage (n=8) and Ⅳ stage (n=4).Of them,three cases were in company with aspiration pneumonia,four cases with infective pneumonia,five cases with air leak and two cases with intra-alveolar hemorrhage.Typical X-ray features included:the pulmonary lucency decreased,wide-ranging net and grain shadowes of high density,and in companing with a lot of air bronchus sign.Conclusion X-ray plain film,especially computed radiography is the most important imaging method in diagnosis of NRDS.